: The research has two major objectives: 1) to provide clinicians with valid instruments to identify patients at risk for suicide; and 2) to indicate those psychological areas that require intervention in order to prevent suicide. Since these cognitive variables are modifiable or reversible by therapeutic intervention, another objective is to provide a conceptual framework for introducing specific interventions to reduce suicidal risk and, consequently, prevent suicide. The proposed investigation will be directed to: 1) validating previous findings in a much larger sample with a longer follow-up period; 2) determining the interaction of the cognitive factors with the clinical variables; and 3) specifying the most powerful clinical, psychological, past and family history, and demographic predictors. Previous studies of suicidal patients suggested that three cognitive factors were associated with high suicide intent: high hopelessness, low level of self-esteem, and impaired problem-solving ability. Other previous studies have supported the role of the first two variables as predictors of eventual suicide. In addition, an association of severity of depression, previous suicide attempts, suicide intent at its worst point, and a history of alcoholism with eventual suicide in an outpatient sample has been found. This investigation will attempt to validate these findings. Data already collected from 1978-1990 on approximately 4,000 patients will be prepared for analysis. New clinical, psychological, past history, and demographic data on approximately 3,000 patients will be collected from patients at two outpatient clinics. Follow-ups will be conducted with these 7,000 patients to determine which patients committed suicide. These data will be analyzed to answer numerous questions, including; 1) Which sets of variables predict eventual suicide? and 2) What reversible or remediable factors predict suicide? Based on these findings, the investigators aim to develop a profile or algorithm for identifying high risk suicidal patients. Further, several self-report measures that can be used quickly and economically to measure suicidal risk will be provided.